Mutation and expression analysis of human BUB1 and BUB1B in aneuploid breast cancer cell lines

Cancer Letters 152 (2000) 193±199

www.elsevier.com/locate/canlet

Mutation and expression analysis of human BUB1 and BUB1B in aneuploid breast cancer cell lines Kenute A. Myrie a, Melanie J. Percy b, James N. Azim b, Christopher K. Neeley b, Elizabeth M. Petty a,b,* a b

Department of Human Genetics, University of Michigan Medical School, Ann Arbor, MI 48109-0638, USA Department of Internal Medicine, University of Michigan Medical School, Ann Arbor, MI 48109-0638, USA Received 20 October 1999; received in revised form 20 December 1999; accepted 21 December 1999

Abstract Genetic instability is a hallmark feature of breast, colorectal and other types of cancers. One type characterized by chromosomal instability is thought to be important in the pathogenesis of many solid tumors displaying aneuploidy. Two related protein kinases and homologues of the yeast checkpoint genes, hBUB1 and hBUB1B, have been implicated in the pathogenesis of colorectal cancers. Mutations in hBUB1 have demonstrated a dominant negative effect by disrupting the mitotic checkpoint when transfected into euploid colon cancer cell lines. In Brca2 de®cient murine cells, Bub1 mutants potentiate growth and cellular transformation. This would suggest that aneuploidy in solid tumors including breast, could be the result of defects in mitotic checkpoint genes and may be responsible for a chromosomal instability phenotype contributing to tumor progression. We conducted mutational analysis of 19 aneuploid breast cancer cell lines. No mutations were found but we identi®ed nine sequence variations including ®ve previously unreported sequence variants in hBUB1B, two of which affect restrictions sites. None of these nucleotide changes predict signi®cant changes in the predicted protein structure. Expression analysis by Northern blot of breast cell lines showed variable expression of hBUB1 and hBUB1B genes. This suggest that while regulation of expression of these genes may be important in cancer, the lack of putative deleterious mutations in the coding sequence does not support a frequent role for mutant hBUB1 and hBUB1B alleles in the pathogenesis of breast cancer. q 2000 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Aneuploidy; Chromosomal instability; Mitotic checkpoint; Breast cancer; hBUB1; hBUB1B

1. Introduction Aneuploidy, a phenotype associated with the gain or loss of whole chromosomes, is found in many types of tumor [1]. It can arise when there is a defect in the mitotic spindle checkpoint, the process ensures that * Corresponding author. Department of Medicine, 4301 MSRB III, 1150 W. Medical Center Drive, Ann Arbour, MI 48109-0638, USA. Tel.: 11-734-764-1549; fax: 11-734-647-7979. E-mail address: [email protected] (E.M. Petty)

sister chromatids are equally divided upon duplication of the cell. The process of appropriate chromosome alignment along the metaphase plate relies on interactions between the kinetochore and microtubules of the spindle that allow movement of chromosomes to the metaphase plate [2]. Accurate cell division relies on a delay at the mitotic checkpoint before the cell moves from metaphase to anaphase. Careful monitoring of the kinetochore by mitotic spindle checkpoint genes accounts for the accurate segregation of chromosomes during anaphase. Defects in checkpoint

0304-3835/00/$ - see front matter q 2000 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0304-383 5(00)00340-2

194

K.A. Myrie et al. / Cancer Letters 152 (2000) 193±199

genes have been shown to be a possible mechanism for aneuploidy and the promotion of tumor formation in various tissues [3,4]. A family of evolutionarily conserved genes involved in the mitotic checkpoint and originally identi®ed in yeast has led to the isolation of human homologues. The normal function of the yeast genes, MAD (MAD1, MAD2, MAD), BUB (BUB1, BUB2, BUB3) and MPS1 [5±9], is to arrest the cell during mitosis in response to defects in the spindle or centromere. Two related human BUB1 like-kinases, hBUB1 and hBUB1B (also known as hBUBR1/MAD3-L, but referred to here as hBUB1B in accordance with HUGO nomenclature guide) were found to carry mutations in some colon cancer cell lines [3]. These genes associate with the kinetochore proteins CENPE and CENP-F, which are involved in kinetochore motor function during chromosome alignment and early kinetochore assembly, respectively [10]. hBUB1 and hBUB1B are protein kinases that contain two conserved domains (CD1 and CD2), which are involved in directing kinetochore localization and binding other related BUB proteins as well as having a kinase domain for phosphorylation function [3]. These two domains are present in yeast, mouse and human. The human and mouse homologues contains an additional nuclear localization signal that is absent in yeast [3]. Colorectal cell lines when mutated in hBUB1 exhibited chromosomal instability and aneuploidy and fail to arrest in mitosis when treated with drugs that disrupt microtubule movement [3]. Subsequent transfection studies of the mutant alleles to euploid cells disrupt the mitotic checkpoint suggesting that mutations in the hBUB1 gene have a dominant negative effect. Breast cancer contributes to a high percentage of morbidity and mortality among women. The molecular basis of breast cancer is complex, involving multiple genetic, epigenetic and environmental factors. While only 5±10% of breast cancer is clearly hereditary most often due to BRCA1 or BRCA2 germline mutations [11], multiple somatic clonal genetic events deregulating normal growth and cell cycle checkpoints are thought to contribute to the more common sporadic disease. In approximately 10±15% of colon cancers alterations in mismatch and DNA damage repair genes result in genetic instability and accumulation of mutations [12]. In contrast, breast tumorigen-

esis appears not to be frequently associated with mismatch repair defects but appear to be driven by other undescribed molecular genetic events associated with aneuploidy so widely observed in breast tumors. One mechanism associated with aneuploidy is chromosomal instability. Recent analysis of thymic lymphomas from Brca2 de®cient mice suggested that, inactivating mutations in mitotic checkpoint genes Bub1-3 and Mad1-3 work synergisticly with alterations in Brca2 to cause structural abnormalities in chromosomes [13]. This synergistic mechanism may be responsible for sustaining and promoting cell division, thereby bypassing the mitotic checkpoint resulting in aneuploidy and contribute to transformation and tumorigenesis. This study also demonstrated that dominant negative mutations in hBUB1 have the capacity to reverse growth arrest in Brca2 de®cient cells and initiate neoplastic transformation. These ®ndings further implicate the importance of alterations in mitotic arrest genes with regard to breast cancer and possible treatment with drugs that target the mitotic spindle or checkpoint genes, warranted further investigation of hBUB1 and hBUB1B in breast cancer. Given the role of these genes in colorectal cancers, aneuploidy seen in breast cancers and most solid tumors, we hypothesize that defects in the mitotic arrest genes hBUB1 and hBUB1B may contribute to chromosomal instability in breast tumors and the aneuploid phenotype associated with these cancers. We, therefore, investigated the role of these genes in breast cancer by screening a panel of 19 breast tumor cell lines.

2. Materials and methods 2.1. Cell culture The cells used for this study were acquired from two sources. Ten cytogeneticly characterized SUM breast cancer cell line and four human papilloma virus immortalized non-tumorigenic mammary cell lines (HPV) were developed at the University of Michigan Comprehensive Cancer Center [14±18]. The growth conditions used for these cell lines are available at http://p53.cancer.med.umich.edu/clines/ clines.html. All other breast cell lines were acquired

K.A. Myrie et al. / Cancer Letters 152 (2000) 193±199

from the American Type Culture Collections (Rockville, MD) and grown according to ATCC cell culture protocols (www.atcc.org). 2.2. RNA isolation and Northern blot analysis Total RNA was isolated from cell lines by lysing using Trizol reagent according to manufacturer's instructions (Gibco BRL) and resuspended in ultrapure DNase, RNase-free water (Gibco BRL). 10 mg of total RNA was fractionated on a 1.5% denaturing agarose gel and transferred to Hybond XL membrane (Amersham Life Science) in 10 £ SSC buffer solution. Hybridization was performed in 5 £ SSC, 5 £ Denhardts at 608C overnight using hBUB1 or hBUB1B full-length RT-PCR products for probe labeled by random primer. The control probe was a 32P-labeled 407 bp cDNA GAPDH PCR product (5 0 -GGGAGCCAAAAGGGTCATCA-3 0 ; 5 0 -TTTCTAGACGGCAGGTCAGGT3 0 ). After hybridization the blots were washed twice at 608C in 2 £ SSC, 0.1% SDS and exposed on Kodak XAR-5 ®lm. 2.3. DNA isolation and Southern blot analysis Genomic DNA was isolated from breast cell lines by treatment with 200 mg/ml of proteinase K, 1% SDS, TNE (500 mM Tris, 20 mM EDTA, 10 mM NaCl, (pH 9.0)) extraction with 1:1 phenol/chloroform. Ethanol precipitated DNA was resuspended in 1 £ TE. 5 mg of DNA were digested to completion with excess restriction endonuclease MspI, EcoRI, TaqI (Gibco BRL) under reaction condition. Digested DNA was electrophoresed in a 0.9% agarose gel in 1 £ TBE and transferred to a Hybond N1 membrane (Amersham Life Science) in 20 £ SSC. Subsequently the blot was hybridized in 1 mM EDTA, 0.5 NaH2PO4, 0.7% SDS while the probe was blocked with 50 mg human cot1-DNA at 608C. Hybridization was performed overnight with 32P-labeled hBUB1 (3.7 kb) and hBUB1B (3.3 kb) full-length cDNA probes. After hybridization blots were washed in 2 £ SSC, 0.1% SDS at 608C then exposed to Kodak XAR-5 ®lm. 2.4. Mutation screening RT-PCR was performed with 5 mg of total RNA using the Superscript Preampli®cation System (Gibco

195

BRL) and oligo dT primers. hBUB1 (GenBank accession no. AF046078) and hBUB1B (GenBank accession no. AF046079) transcripts were ampli®ed using three sets of overlapping primers for each gene. The primers were previously published [3]. PCR products were puri®ed using the High Pure PCR Product Puri®cation Kit (Boehringer Mannheim) and quanti®ed with ¯uorometer or DNA mass ladder (Gibco BRL). Sequencing was performed using the Thermosequenase Cycle Sequencing Kit (Amersham Life Science) with published primers [3] and internal primers generously provided by D. Cahill according to the manufacturer. Reaction conditions consisted of 30 cycles at 948C for 30 s, 588C for 30 s, 728C for 1 min. Samples were electrophoresed on 6% polyacrylamide gel consisting of 10 £ glycerol tolerant buffer, (USB), urea, Long Ranger 50% gel solution (FMC Bioproducts), TEMED and 10% APS at 75 W. Dried gels were exposed to Kodak XAR-5 ®lm at 2708C for 24± 72 h. 2.5. Densitometry analysis The mRNA levels for each of the cell lines were analyzed by densitometry (Alpha Innotech IS-1000 Digital Imaging System, version 2.00) according to the manufacturer instructions. BUB1 and BUB1B levels were normalized to GAPDH levels by calculating the ratio of BUB1 or BUB1B to GAPDH. Ratios between cancer cell lines and HPV immortalized nontumorigenic lines were compared. 3. Results and discussion Both genes were expressed in all cell lines and one major band of expected size 3689 bp for hBUB1 or 3309 bp for hBUB1B was observed from RT-PCR products (Fig. 1). Northern blot analysis was used to grossly assess the transcripts of the BUB1 and hBUB1B genes. The hBUB1B gene appeared to show low level of expression in SUM 185, SUM 225, SUM 1315 and a HS578 (Fig. 1). By visual observation a relative increase in expression was detected in MCF7. The hBUB1 transcript expression appeared reduced in HS578 and was moderately increased in MCF7 (Fig. 1). To con®rm these differential levels of expression detected densitometry was performed. The two HPV immortalized non-tumori-

196

K.A. Myrie et al. / Cancer Letters 152 (2000) 193±199

Fig. 1. mRNA expression analysis of hBUB1 and hBUB1B in 16 aneuploid breast tumor cell lines, one near diploid breast tumor line and two immortalized breast lines. The GADPH gel is shown as a loading control.

genic lines gave similar ratios to GAPDH and were averaged and use to compare with other cell lines. In HS578, SUM 185 and SUM1315 lines BUB1B expression was approximately 2.4 times lower in intensity while in SUM 225 it was slightly less as compared to HPV immortalized non-tumorigenic lines. MCF 7 demonstrated 8.3 times greater intensity as compared to immortalized non-tumorigenic HPV lines. For BUB1, signal intensity of the transcript was 2.0 times lower in HS578 while in MCF7 BUB1 was 9.3 times more intensely expressed than HPV immortalized-non-tumorigenic cells. It is dif®cult to accurately assess the importance of variable expression given that the samples analyzed were cancer cell lines and matched normal samples were not available for analysis. The levels of transcript expression may vary depending on a variety of factors including the rate of cell growth at the time the cells were harvested for RNA isolation. To minimize this potential for variability, cells were harvested at 80% con¯uence. Similarly, the phosphorylation states of BUB1 and BUB1B RNA changes at different times during the cell cycle. Southern blot analysis (data not shown) did not reveal any structural changes that could explain the

decreased expression seen in HS578 or the three SUM cell lines. Neither could it account for apparent increased expression in the MCF7 cell line. Therefore, we sequenced whole transcript of the hBUB1 and hBUB1B genes to see if alterations in the coding sequence could account for the variability of expression. Mutational analysis of the hBUB1 transcripts in 19 aneuploid breast cell lines revealed no changes in the coding regions of this gene (data not shown). Perhaps in breast cancers the hBUB1 is not often mutated within its coding region but could be deregulated by other mechanisms such as promoter methylation. In contrast, sequencing the transcript of the hBUB1B gene from RT-PCR products identi®ed ®ve new sequence variants (Table 1) that are most likely polymorphism based on their lack of predicted amino acid change. Four previously identi®ed polymorphisms within the coding regions were also identi®ed [3]. To date, these are the ®rst sequence variants of the mitotic checkpoint gene, hBUB1B, to be identi®ed in breast tumors. Two of the ®ve new sequence variants are responsible for restriction site changes. An A ! G change at nucleotide position 3141 creates a new restriction site recognized by six isoschizomers with

K.A. Myrie et al. / Cancer Letters 152 (2000) 193±199

197

Table 1 hBUB1B sequence variations Nucleotide position

Codon

324 1088 1206 2899 a 3141 a

AAA(Lys) ! AAG(Lys) CAA(Gln) ! CGA(Arg) GCG(Ala) ! GCA(Ala) CTG(Leu) ! TTG(Leu) AAA(Lys) ! AAG(Lys)

3298/3299 a 3301 a 3306

GCA ! TGA (alt.stop) CTA ! ATA (3 0 UTR) AAA ! AAG (3 0 UTR)

a

Restriction site changes

Agg/cct, cctnn/nnnagg, new site created; EcoNI, Pme55I, StuI, AatI,Eco147I, SseBI Caynn/nnrtg, MslI site deleted

Sequence variants that have been identi®ed for the ®rst time. The bases underlined were changed in breast samples.

only two unique sites (Table 1). This new restriction site within the coding region of the gene is approximately 54 bases from the stop codon and lies outside the CD2 domain. This sequence variation does not affect any exon/intron boundaries and would, therefore, be predicted to have no effect on mRNA expression. Two other heterozygous changes at nucleotide positions 3298 G ! T and 3299 C ! G in the 3 0 UTR destroyed a Msl I site and result in another stop codon (GCA ! TGA). Based on location, it is unlikely that this stop codon has any effect on gene expression, in particular, no alternate stop codon usage. Furthermore our expression and sequencing data of cDNA revealed only one transcript and, therefore, does not appear to be affected by the new stop codon in the 3 0 untranslated region. The other two new variants, 3301 C ! A (Leu to Ile) and 3306 A ! G (Lys to Lys) did not affect any restriction sites when checked against the Baylor College of Medicine Search Launcher:sequence utilities restriction site (http://dot.imgen.bcm.tmc.edu:9331/seq-util). The polymorphisms that are responsible for deleting or creating new restriction sites may be useful for population studies involving cancer and chromosome instability genes. In particular, allele association studies with aneuploid tumors may help in identifying at risk individuals for breast cancers and allow for early clinical or genetic counseling intervention. Within the coding region of hBUB1B we identi®ed a sequence difference from that previously published [3] at (Accession number AF046079, GenBank) in all our cell lines. The published sequence reported an A at nucleotide 355 that was absent from our sequencing

data. However, we found an A at nucleotide position 382 in the hBUB1B gene sequence of all breast cell lines. Our sequence of the hBUB1B gene was con®rmed by double pass sequencing of independent cDNA samples. The difference in sequence results in a change in the amino acid sequence beginning at nucleotide 355±382 from the published Ser-CysArg-Ser-Thr-Thr-Arg-Arg-Lys [3], to Ala-Val-GluAla-Leu-Gln-Gly-Glu-Lys and occurs in the highly conserved CD1 domain, one of three functional domains in the hBUB1B gene. The domain is important for binding to BUB3 and it shares homology with BUB1 and MAD3-like proteins as well as function in kinetochore localization at prometaphase [19]. BUB3 is required for the kinetochore localization of BUB1 in response to unattached kinetochores and thereby activate the mitotic checkpoint. In so doing cdc 20 is inhibited and the APC (anaphase promoting complex) is prevented from being activated. The amino acid substitution that occurs as a result of the sequence difference indicates that there are four more hydrophobic amino acids in this domain than previously reported. A new version of the hBUB1B sequence has since been reported to GenBank (version gi 5705870) by D. Cahill (personal communication). Currently no functional capabilities are associated with the 3 0 UTR of the hBUB1B gene. However, the 3 0 UTR may contain yet to be identi®ed binding sites for nuclear factors that may be involve in hBUB1B regulation. In such a case the 3 0 UTR sequence variants may have an effect on hBUB1B function and explain the lowered gene expression seen in the SUM lines and HS578. Published data suggests that

198

K.A. Myrie et al. / Cancer Letters 152 (2000) 193±199

BUB1 and BUB1B kinase activity and localization changes throughout the cell cycle [20,21], therefore, regulation of BUB1 or BUB1B function may be affected by post-translational modi®cation and protein levels. Certainly, these changes if found to be signi®cant in further studies could have variable effects on different populations, age, diet and environment. It is possible that the observed decrease in mRNA levels in HS578 and SUM cell lines may be of no consequence and the cells in actuality may have similar protein levels. Continued study of the BUB1 and BUB1B genes will be enhanced by future determination of expressed protein levels compared to mRNA levels. If the observed differences in expression turn out to be real it will be challenging to reconcile how polymorphism may contribute to aneuploidy. Possible mechanisms that may contribute to the variable expression may include hypermethylation at promoter sites or within the 3 0 UTR. It has been demonstrated that hypermethylation of promoters is a common mechanism in some cancers whereby genes can become inactivated [22]. Similarly, there might be defects in other related mitotic arrest genes that contributes to chromosomal instability and the aneuploid phenotype that have yet to be studied. Similarly, reduced expression of hBUB1B may affect binding to other protein products in a multi-protein complex involved in controlling the mitotic checkpoint. Poor binding with CENP E or the kinetochore could abrogate the link between the checkpoint and the kinetochore resulting in loss of cdc20 inhibition and activation of the APC allowing cells to proceed through anaphase and becoming aneuploid. Increased expression in MCF7 may cause mitotic delay resulting in aneuploidy. It has been shown previously that overexpression of Mps1, which lies upstream of the BUB and MAD genes alone is suf®cient to cause mitotic delay [23]. Since no other mutations were identi®ed except for those previously mentioned it is also possible that there may be defects in trans-acting factors or cis-sequence elements that regulate these genes and might explain the variable expression seen in these cell lines. Further investigation will be needed to determine which mechanisms could account for the aneuploid phenotype seen in these breast tumor cell lines associated with variable BUB1 and BUB1B expression. Although our data does not shed any new light on the

etiology of breast cancer we feel that the newly described polymorphisms may help in future studies with regards to allele association and disease. Acknowledgements We wish to thank S. Ethier for kindly providing us with SUM cell lines, D. Cahill for helpful comments and providing the hBUB1 and BUB1B primer sequences. This work was sponsored in part by NIH grants K08CA66613-01 and R01CA72877 (E.M.P.). References [1] F. Mitleman, F. Mertens, B. Johansson, breakpoint map of recurrent chromosomal rearrangements in human neoplasm, Nat. Genet. 15 (1997) 417±474. [2] S.A. Jablonski, G.K.T. Chan, C.A. Cooke, W.C. Earnshaw, T.J. Yen, The hBUB1 and hBUBR1 kinase sequentially assemble onto kinetochore during prophase with hBUBR1 concentrating at the kinetochore plates in mitosis, Chromosoma 107 (1998) 386±396. [3] D.P. Cahill, C. Lengauer, J. Yu, G.J. Riggins, J.K.V. Willson, S.D. Markowitz, K.W. Kinzler, B. Vogelstein, Mutations of mitotic checkpoint genes in human cancers, Nature 392 (1998) 300±303. [4] Y. Li, R. Benezra, Identi®cation of a human mitotic checkpoint gene: hsMAD2, Science 274 (1996) 246±248. [5] R. Li, A.W. Murray, Feedback control of mitosis in budding yeast, Cell 66 (1991) 519±531. [6] M.A. Hoyt, L. Totis, B.T. Roberts, S. cerevisiae genes required for cell cycle arrest in response to loss of microtubule functions. Cell 66 (1991) 507±517. [7] Y. Wang, D.J. Burke, Checkpoint genes required to delay cell division in response to nacodazole to impaired kinetochore function in the yeast Saccharomyces cerevisiae, Mol. Cell Biol. 15 (1995) 6838±6844. [8] E. Weiss, M. Winey, The Saccharomyces cerevisiae spindle pole body duplication gene MPS1 is part of a mitotic checkpoint, J. Cell Biol. 132 (1996) 111±123. [9] F. Pangilian, F. Spencer, Abnormal kinetochore structure activates the spindle assembly checkpoint in budding yeast, Mol. Cell Biol. 7 (1996) 1195±1208. [10] G.K.T. Chan, B.T. Schaar, T.J. Yen, Characterization of the kinetochore binding domain of CENP-E reveals interactions with the kinetochore proteins CENP-F and hBUBR1, J. Cell Biol. 143 (1998) 49±63. [11] J.M. Hall, M.K. Lee, B. Newman, J.E. Morrow, L.A. Anderson, B. Huey, M.L. King, Linkage of early onset breast cancer to chromosome 17q21, Science 250 (1990) 1684. [12] C. Lengauer, K.W. Kinzler, B. Vogelstein, Genetic instabilities in human cancers, Nature 396 (1998) 643±649. [13] H. Lee, A.H. Trainer, L.S. Friedman, F.C. Thistlethwaite, M.J.

K.A. Myrie et al. / Cancer Letters 152 (2000) 193±199

[14]

[15]

[16]

[17]

[18]

Evans, B.A.J. Ponder, A.R. Venkitaraman, Mitotic checkpoint inactivation fosters transformation in cells lacking the breast cancer susceptibility gene, Brca2. Mol. Cell 4 (1999) 1± 10. S.P. Ethier, L.M. Mahacek, W.J. Gullick, T.J. Frank, B.L. Weber, Differential isolation of normal luminal mammary epithelial cells and breast cancer cells from primary and metastatic sites using selective media, Cancer Res. 53 (1993) 627± 635. S.P. Ethier, K.E. Kokeny, J.E. Ridings, C.A. Dilts, erbB family receptor expression and growth regulation in a newly isolated human breast cancer cell line, Cancer Res. 56 (1996) 899±907. R. Garcia, C.L. Yu, A. Hudnall, R. Catlett, K.L. Nelson, T. Smithgall, D.J. Fujita, S.P. Ethier, R. Jove, Constitutive activation of Stat3 in ®broblasts transformed by diverse oncoproteins and in breast carcinoma cells, Cell Growth Diff. 8 (1997) 1267±1276. C.I. Sartor, M.L. Dziubinski, C.L. Yu, R. Jove, S.P. Ethier, Role of epidermal growth factor receptor and STAT3 activation in autonomous proliferation of SUM-102PT human breast cancer cells, Cancer Res. 57 (1997) 978±987. K.M. Ignatoski, S.P. Ethier, Constitutive activation of

[19] [20]

[21]

[22]

[23]

199

pp125fak in newly isolated human breast cancer cell lines, Breast Cancer Res. Treat. 54 (2) (1998) 173±182. S.S. Taylor, E. Ha, F. McKeon, The human homologue of Bub3 is required for kinetochore localization of Bub1 and a Mad3/ Bub1-related protein kinase, J. Cell Biol. 142 (1998) 1±11. J. Basu, H. Bousbaa, E. Logarinho, Z. Li, B.C. Williams, C. Lopes, C.E. Sunkel, M.L. Goldberg, Mutations in the essential spindle checkpoint gene bub1 cause chromosome missegregation and fail to block apoptosis in Drosophila. J. Cell Biol., 146 (2) (1999) 13-28. G.K. Chan, S.A. Jablonski, V. Sudakin, J.C. Hittle, T.J. Yen, Human BUBR1 is a mitotic checkpoint kinase that monitors CENP-E functions at kinetochores and binds the cyclosome/ APC. J. Cell Biol., 146 (5) (1999) 941-954. A. Merlo, J.G. Herman, L. Mao, D.J. Lee, E. Gabrielson, P.C. Burger, S.B. Baylin, D. Sidransky, 5 0 CpG island methylation is associated with transcriptional silencing of the tumor suppressor p16/CDKN2/MTS1 in human cancers, Nat. Med. 1 (1995) 686±692. K.G. Hardwick, E. Weiss, F.C. Luca, M. Winey, A.W. Murray, Activation of the budding yeast spindle assembly checkpoint without mitotic spindle disruption, Science 273 (1996) 953±956.